September 2006

The 2006 Elections: Are We Ready?

On September 22, the AEI-Brookings Election Reform Project held a public conference to discuss the current status of election reform efforts. After a keynote address by Representative Rush Holt (D-N.J.), the conference’s first panel brought together technology experts from leading universities and an independent research organization to consider the likelihood of Florida-like debacle in 2006. Following the technology panel, a second panel comprised of public policy experts assessed the progress states have made in complying with the Help America Vote Act (HAVA) of 2002.

Keynote Address

The Honorable Rush Holt (D-N.J.)
U.S. House of Representatives

The answer to the conference’s question of “Are we ready?” is “No.” HAVA encouraged localities to replace obsolescent voting machinery, but in many cases the transition to new technology was rushed and gave rise to voting methods that are unverifiable and unreliable. Additionally, there are several instances where unfamiliarity with new technology has led to administrative errors, as was the case in a recent primary election in Montgomery County, Maryland. Machine vendors should make their software public so as to permit public scrutiny. To be sure, there are areas in which we can praise the transition to electronic voting. At the same time, increased accessibility and accountability are necessary. A voter-verified paper audit trail and random public audits are essential to maintain the legitimacy of elections.

Panel I: Progress and Pitfalls in Voting Technology

Steven Hertzberg
Election Science Institute

There were numerous procedural problems in the 2006 primary election. Although voters generally reported confidence in the new Digital Recording Electronic (DRE) election system, several incidents could be disastrous if they were to be encountered on a larger scale. For example, 87 of the 485 paper audit rolls used were lost after the election. Furthermore, vote tallies from the paper trails did not match the electronic tallies upon comparison.

Beyond conducting additional testing in 2006 and 2007, we should reevaluate our auditing procedures, increase field testing of voting machines, and conduct further analysis of specific problems that commonly arise, such as printer jams.

Michael I. Shamos
Carnegie Mellon University

While it is important not to understate the severity of problems that have arisen in election administration, there is reason to be optimistic about the future.  Many of the flaws in electronic voting machines that do exist are dangerous, but that does not mean they are not easily repaired. Rather, we should remember to approach the problem the way we would approach a technical problem in the airline or banking industries. We should address and fix the problem, not abandon the method.

A side-by-side comparison of the leading voting methods confirms that paperless DRE machines and optical scan ballot markers offer the best balance of usability, secrecy, security, reliability, and accessibility. While this conclusion disappoints the many ardent advocates of paper trails, such trails may hinder a voter’s ability to maintain the secrecy of his or her ballot. They are also not typically accessible to the blind. Anticipating objections on the part of individuals who believe paper trails are necessary in order to provide a check against glitches, fraud, and tampering, it must be noted that security can be ensured by frequent audits both before and after an election.

Charles Stewart III
Massachusetts Institute of Technology

A “big-picture” view of election administration illuminates some important trends as new voting machines have been implemented over the past several election cycles. For example, an examination of the changing percentage of residual votes--the number of improperly counted ballots--in different elections shows that states with high numbers of uncounted votes tend to improve by the next election cycle as administrators address problems, while states with less severe problems tend to display relatively constant levels of residual voting. Additionally, certain types of voting machines, especially DRE machines, are successful in reducing the frequency of improperly marked ballots.

Panel II: HAVA: What Has Been Done? What Remains to Do?

The Honorable Deborah Markowitz
Vermont Secretary of State

Vermont is proud to use election practices that are “trailing so far behind that they are leading the pack.” In the majority of its precincts, paper ballots are counted by hand, while ballots in only a few counties are scanned by computer. In order to aid disabled voters, Vermont has a popular and reliable vote-by-phone system. Vermont also uses an informal system of bipartisan checks and balances in which at least one person from each party is present during the counting of votes, the guarding of ballot boxes and bags, and the sealing of records.

New equipment and statutes will inevitably cause more lawsuits after the November election. Lawsuits over provisional ballots are particularly interesting. In Vermont, voter registration checklists are going to make a big difference. Seven thousand five hundred voters were removed from voter rolls recently because they were registered in more than one location. We can expect litigation to arise due to confusion about identification requirements for voter registration. Such litigation, however, is healthy for democracy.

Donetta Davidson
Election Assistance Commission (EAC)

The EAC distributes various voting resources to states, including a quick-start management guide to outline various plans to program and fix voting equipment, guidelines on how to print and distribute ballots properly, and information on voting system security.

Technology problems will exist during the November elections, but mitigating procedures must be established. For example, during the Rhode Island primaries, special voting equipment for the disabled was not prepared because election officials did not feel enough voters would need it. Delays resulted, as the machines needed time to be set up. It was an irresponsible procedure for using funds for special technology.

Election administrators should work hard to encourage experienced individuals--academics and the like--to volunteer at polling places in order to ensure that elections run smoothly and that the public has confidence in the outcome.

Edward Foley
Ohio State University

A very close election in November could combine with delays in counting disputed votes to cause a crisis in January. Election problems are not the only trigger of prolonged litigation. A close race and a team of lawyers are sufficient. Thus, though there will probably be technological failures across the country, even technological perfection would not eradicate the potential for litigation, especially over absentee and provisional voting.

Cuyahoga County, Ohio, is now urging voters to use absentee ballots to reduce pressure on voting machines and poll workers. But absentee voting carries increased risk of fraud, opening doors for lawsuits. Such litigation is very easy to initiate but takes a long time to run its course.

Also, provisional voting is risky, as it is necessary to check provisional ballots against registration lists. The case with the greatest potential for litigation, however, is that in which a voter is turned away in the morning and cannot return at night, a foreseeable result of a lack of available provisional ballots at the polls.

It seems that we need to be prepared for a post-Election Day environment in which voters have lost confidence and trust in the electoral system.

Doug Lewis
The Election Center

Eighty-two million voters will be voting on new equipment in November, and it is impossible to predict the problems that will arise. This makes it difficult for election officials to prevent controversy on Election Day.

With every major change in election administration, it takes about three elections to establish sufficient procedures to prevent significant controversy. The timeframe expected for proper implementation of new technology and procedures may thus be unrealistic. It takes time to absorb major changes, and further changes made in panic might only make matters worse in November.

A major problem is that many people simply do not understand that it is difficult to recruit poll workers in urban areas. Furthermore, poll workers bear unrealistic expectations of perfection. It is inevitable that poll workers will make mistakes, but there exists to an extent an unfair belief that human mistakes with regard to elections ought to result in severe punishment or dismissal.

AEI research assistant Timothy J. Ryan and intern Seth Rokosky prepared this summary.